Title: Probabilistic solution discovery algorithm for the orienteering problem
Author: Jose Emmanuel Ramirez-Marquez, Sadan Kulturel-Konak, Claudio M. Rocco Sanseverino
School of Systems and Enterprises, Stevens Institute of Technology, Castle Point on Hudson, Hoboken, NJ 07030, USA.
Management Information Systems, Penn State Berks, Tulpehocken Road, P.O. Box 7009, Reading, PA 19610, USA.
Facultad de Ingeniería, Universidad Central de Venezuela, Caracas, Venezuela
Abstract: Orienteering is a competition where participants have a specific time to travel between initial and final locations. This competition is known as the orienteering problem (OP) where the objective is to maximise the reward of visiting sites while not violating constraints on time and final location. Although very good solutions can be obtained via search space heuristics, their coding, development and implementation still require extensive familiarisation and background on the technique of choice. This paper presents an evolutionary algorithm that offers simple and efficient analysis for the OP via three interrelated steps: solution generation via Monte Carlo (MC) simulation; analysis and computation of the reward and distance of these solutions and an evolutionary technique that drives the selection of potentially optimal solutions. The algorithm is tested against benchmark techniques for problems previously solved in the literature and newly developed larger size test problems.
Keywords: orienteering problem; optimisation; probabilistic solutions; discovery algorithms; reward maximisation; time constraints; search space heuristics; coding; evolutionary algorithms; solution generation; Monte Carlo simulations; optimal solutions; benchmarks; benchmarking; large size test problems; computation; distance; navigational skills; outdoor competitions.
Int. J. of Industrial and Systems Engineering, 2010 Vol.6, No.1, pp.45 - 61
Available online: 06 Jul 2010